OBJECTIVE: To determine the frequency and possible origins of calibration errors of the Goldmann applanation tonometer and to investigate intraobserver and interobserver agreement in the measurement of such calibration errors. DESIGN: Evaluation of diagnostic test or technology. PARTICIPANTS: One hundred thirty-two slit-lamp-mounted Haag-Streit Goldmann tonometers (Model AT 900 C/M; Haag-Streit, Bern, Switzerland). METHODS: Calibration error testing was performed using the standard calibration error check weight bar. A single observer tested 25 instruments on 2 consecutive days. Two observers independently measured calibration errors of another 40 instruments. A single observer performed prospective testing of the remaining Haag-Streit Goldmann tonometers, 6 months after their annual calibration by the bioengineering department of the hospital. Four months later, the same observer retested the instruments that were found to be faulty earlier. MAIN OUTCOME MEASURE: Goldmann applanation tonometer calibration error. RESULTS: Only 4% of Goldmann tonometers were found to be within the manufacturer's recommended calibration error tolerance (+/-0.5 mmHg) at 20 mmHg. Twenty-eight percent of instruments had calibration errors of more than +/-2 mmHg at the 20-mmHg testing level. An additional 12.12% had calibration errors of more than +/-2 mmHg at the 0- or 60-mmHg testing levels, or both. Four months after the initial testing and repair of the 53 faulty instruments, 20.75% again were found to have calibration errors of more than +/-2 mmHg at any one or more testing level(s). Intraclass correlation coefficients for intraobserver agreement at 20 mmHg were 0.57 (95% confidence interval [CI], 0.25-0.78) for positive error, 0.82 (95% CI, 0.65-0.92) for negative error, and for interobserver agreement at 20 mmHg, 0.83 (95% CI, 0.70-0.90) for positive error and 0.83 (95% CI, 0.70-0.90) for negative error. CONCLUSIONS: The authors found good agreement in measurement of Goldmann applanation tonometer calibration error. A significant number of instruments had clinically unacceptable calibration errors. A more frequent calibration error testing and automation of calibration error check is desirable. FINANCIAL DISCLOSURE(S): The authors have no proprietary or commercial interest in any materials discussed in this article.
OBJECTIVE: To determine the frequency and possible origins of calibration errors of the Goldmann applanation tonometer and to investigate intraobserver and interobserver agreement in the measurement of such calibration errors. DESIGN: Evaluation of diagnostic test or technology. PARTICIPANTS: One hundred thirty-two slit-lamp-mounted Haag-Streit Goldmann tonometers (Model AT 900 C/M; Haag-Streit, Bern, Switzerland). METHODS: Calibration error testing was performed using the standard calibration error check weight bar. A single observer tested 25 instruments on 2 consecutive days. Two observers independently measured calibration errors of another 40 instruments. A single observer performed prospective testing of the remaining Haag-Streit Goldmann tonometers, 6 months after their annual calibration by the bioengineering department of the hospital. Four months later, the same observer retested the instruments that were found to be faulty earlier. MAIN OUTCOME MEASURE: Goldmann applanation tonometer calibration error. RESULTS: Only 4% of Goldmann tonometers were found to be within the manufacturer's recommended calibration error tolerance (+/-0.5 mmHg) at 20 mmHg. Twenty-eight percent of instruments had calibration errors of more than +/-2 mmHg at the 20-mmHg testing level. An additional 12.12% had calibration errors of more than +/-2 mmHg at the 0- or 60-mmHg testing levels, or both. Four months after the initial testing and repair of the 53 faulty instruments, 20.75% again were found to have calibration errors of more than +/-2 mmHg at any one or more testing level(s). Intraclass correlation coefficients for intraobserver agreement at 20 mmHg were 0.57 (95% confidence interval [CI], 0.25-0.78) for positive error, 0.82 (95% CI, 0.65-0.92) for negative error, and for interobserver agreement at 20 mmHg, 0.83 (95% CI, 0.70-0.90) for positive error and 0.83 (95% CI, 0.70-0.90) for negative error. CONCLUSIONS: The authors found good agreement in measurement of Goldmann applanation tonometer calibration error. A significant number of instruments had clinically unacceptable calibration errors. A more frequent calibration error testing and automation of calibration error check is desirable. FINANCIAL DISCLOSURE(S): The authors have no proprietary or commercial interest in any materials discussed in this article.